Generally, in case of a media signal, an encoder compresses a multi-channel signal into a mono- or stereo-type downmix signal instead of compressing each multi-channel signal. The encoder then transfers the compressed downmix signal and spatial information or extension data to a decoder or stores them in a storage medium. And, the decoder reconstructs original multi-channels using the compressed downmix signal and the spatial information.
The number of channels, which can be basically compressed and reconstructed by encoder and decoder, is preset. In N-M-N channel configuration, on the assumption that a front ‘N’ is the number of channels to be transferred by an encoder, that ‘M’ is the number of compressed downmix signals, and that a rear ‘N’ is the number of channels to be reconstructed by a decoder, the encoder and decoder basically provide 5-1-5 channel configuration, 5-2-5 channel configuration, 7-2-7 channel configuration, 7-5-7 channel configuration, etc.
In case of the number of channels less than a channel configuration supported by an encoder, the channels are mapped to a channel structure supported by the encoder and then encoded. In particular, in case that channels less than the channels supported by an encoder are inputted to the encoder, encoding is carried out on the assumption that channels amounting to a difference between the number of channels compressible by the encoder and the number of channels inputted to the encoder have a virtual value. In this case, the encoder generates spatial information required for a decoder to reconstruct the channels having the virtual value and then transfers the generated spatial information to the decoder.